# Free Token Counter

> Free, privacy-first token counter and AI API cost calculator at https://freetokencounter.com covering 650+ models. Counts tokens for OpenAI (exact, in-browser tokenizer), Claude, Gemini, DeepSeek, Mistral, Grok, Llama, Qwen, and hundreds more, estimates request and monthly API cost, and tracks context-window usage. Prompt text never leaves the browser.

## Key facts

- One English word is roughly 1.33 tokens; 1,000 tokens is about 750 English words or 1.5 pages.
- Token counts are exact for OpenAI (o200k_base/cl100k_base), Llama 3, Qwen 3, and DeepSeek V3/R1 models - their tokenizers run in the browser. Claude and Gemini counts are labeled estimates because those tokenizers are not public.
- Flagship model pricing is verified daily against official provider pricing pages; 600+ long-tail models use community-tracked pricing, clearly labeled.
- All pricing data is free to reuse with attribution via https://freetokencounter.com/api/pricing.json (CORS enabled).
- Price changes are detected automatically and published at https://freetokencounter.com/pricing-changes.

## Tools

- [Token counter and cost analyzer](https://freetokencounter.com/): paste a prompt, pick a model, see tokens, cost, context usage, optimization suggestions, and cross-provider comparison.
- [Model token counter directory](https://freetokencounter.com/models): dedicated counter pages for every supported model with current pricing.
- [LLM pricing comparison](https://freetokencounter.com/llm-pricing): input/output price per 1M tokens and context windows for all catalog models.
- [Image token calculator](https://freetokencounter.com/image-token-calculator): vision token math for OpenAI, Claude, and Gemini image inputs.
- [Chat token counter](https://freetokencounter.com/chat-token-counter): conversation-level counting with per-message API overhead.
- [Context windows compared](https://freetokencounter.com/context-windows): every model's context window in tokens, words, and pages.
- [Prompt caching calculator](https://freetokencounter.com/prompt-caching-calculator): cached vs uncached input cost modeling.
- [Embeddable widget](https://freetokencounter.com/embed): free iframe token counter for any website.
- [Exact counts with your API key](https://freetokencounter.com/exact-token-count): billing-exact Claude and Gemini counts via official endpoints.
- [Tokens to words converter](https://freetokencounter.com/tokens-to-words) and [words to tokens converter](https://freetokencounter.com/words-to-tokens).
- [How many tokens is X](https://freetokencounter.com/how-many-tokens): direct answers for a word, sentence, page, essay, book, tweet, and more.

## Provider counters

- [OpenAI token counter](https://freetokencounter.com/openai-token-counter) (exact tokenization)
- [ChatGPT token counter](https://freetokencounter.com/chatgpt-token-counter)
- [Claude token counter](https://freetokencounter.com/claude-token-counter)
- [Gemini token counter](https://freetokencounter.com/gemini-token-counter)
- [Llama token counter](https://freetokencounter.com/llama-token-counter)
- [Mistral token counter](https://freetokencounter.com/mistral-token-counter)
- [DeepSeek token counter](https://freetokencounter.com/deepseek-token-counter)
- [Grok token counter](https://freetokencounter.com/grok-token-counter)
- [Qwen token counter](https://freetokencounter.com/qwen-token-counter)

## Data

- [Public pricing JSON](https://freetokencounter.com/api/pricing.json): current model pricing dataset, free to use with attribution.
- [Pricing change tracker](https://freetokencounter.com/pricing-changes): daily changelog of LLM API price changes with [RSS](https://freetokencounter.com/pricing-changes.xml).